Zanzalintinib

Zanzalintinib (XL092) is Exelixis's attempt to take everything cabozantinib does and make it cleaner - same multi-kinase strategy hitting VEGFR2, MET, AXL, and MER, but engineered with a shorter terminal half-life (~16-22h vs cabozantinib's ~99h) to enable faster dose holds and recovery during toxicity events [1][13]. Phase 3 readouts are stacking up across four solid tumor settings: refractory MSS colorectal cancer (STELLAR-303), non-clear cell renal cell carcinoma (STELLAR-304), head and neck squamous cell carcinoma (STELLAR-305), and previously treated neuroendocrine tumors versus everolimus (NCT06943755) [1][3][4][5]. STELLAR-303 published in Lancet in October 2025 - median OS 10.9 vs 9.4 months, HR 0.80 (95% CI 0.69-0.93), p=0.0045 - the first major late-stage win and the data point that turns this from interesting follow-on into a real franchise candidate [1]. The FDA accepted the zanzalintinib + atezolizumab NDA for refractory mCRC under standard review with a PDUFA target action date of December 3, 2026 [14]. Exelixis booked $2.2B in 2024 revenue, almost entirely from cabozantinib (Cabometyx) [6]. The Cabometyx US composition-of-matter patent (7,579,473) expires August 14, 2026, with later formulation/salt and polymorph patents extending protection until generic entry under settled licenses to Teva and Cipla beginning January 1, 2031 [15]. Zanzalintinib is the lifecycle bet against that hard 2031 cliff. If two of the four Phase 3 programs land, Exelixis extends its oncology franchise into the 2030s. If they don't, the company is a single-product story heading into a defined cliff.

Novel compound, never approved anywhere. Four Phase 3 programs are running in parallel - unusual breadth for a single oral oncology asset and a clear signal that Exelixis is throwing money at the lifecycle problem before cabozantinib starts to erode. The FDA accepted the zanzalintinib + atezolizumab NDA for previously treated metastatic colorectal cancer under standard review (not priority review) with a PDUFA date of December 3, 2026 [14]. As of public filings through Q1 2026, no FDA Breakthrough Therapy or Fast Track designations for zanzalintinib have been confirmed in Exelixis press releases or trial registrations; the recent NDA acceptance was specifically standard review [14]. STELLAR-303 (zanzalintinib + atezolizumab vs regorafenib in refractory MSS colorectal cancer) is the most mature: published in Lancet in October 2025 by Hecht et al., positive on overall survival with median OS 10.9 vs 9.4 months and HR 0.80 [1]. STELLAR-304 (zanzalintinib + nivolumab in non-clear cell RCC) and STELLAR-305 (zanzalintinib + pembrolizumab in HNSCC) are both fully enrolled or close to it, with readouts expected in 2026-2027 [3][4]. The NET trial versus everolimus (NCT06943755) was the latest Phase 3 launch and is still recruiting. Merck is co-developing in clear cell RCC through LITESPARK-034 (NCT07489495), combining zanzalintinib with belzutifan - that combination has Merck's full weight behind it and is a separate commercial vector from the Exelixis-led programs [7]. Exelixis filed its FY2025 10-K on February 10, 2026 [6].

Zanzalintinib blocks four receptor tyrosine kinases at once. VEGFR2 is the receptor on blood vessel cells that tumors hijack to grow their own blood supply - block it and you starve the tumor. MET is a receptor that tells cancer cells to migrate and invade; it's also the bypass kinase that gets amplified when tumors escape EGFR inhibitors or develop metastatic potential [8]. AXL and MER (MERTK) operate on two compartments at once, which is the part most reviewers underweight: (1) on tumor cells themselves, AXL drives epithelial-to-mesenchymal transition, motility, and resistance to targeted therapy and chemotherapy - direct AXL inhibition blunts invasion and reduces drug-resistance signaling; (2) on tumor-associated macrophages and dendritic cells, AXL/MERTK mediate efferocytosis (clearance of apoptotic tumor cells) and the immunosuppressive 'M2-like' macrophage polarization that protects tumors from T cell attack. Blocking AXL/MERTK on the myeloid compartment disrupts both signals, shifting the microenvironment toward a more immunogenic state. Both arms - direct anti-tumor and immune re-education - converge on the same therapeutic hypothesis: an inhibitor that hits VEGFR2 + AXL/MER turns a checkpoint-resistant 'cold' tumor into one that responds to PD-1/PD-L1 blockade. This is why zanzalintinib is consistently paired with checkpoint antibodies in Phase 3 and is the same logic that drove cabozantinib + nivolumab to approval in RCC [1][9]. On the pharmacokinetic side: cabozantinib's ~99-hour terminal half-life means a dose reduction takes days to wash out during a toxicity event, which forced the well-known real-world 60→40→20mg staircase. Zanzalintinib's terminal half-life of 16-22 hours from Phase 1 dosing supports once-daily administration while allowing clinicians to hold the drug and resume within ~24-48 hours of a tolerability event [13]. The therapeutic argument is that faster recovery enables more time at the target dose rather than chronic dose attenuation - historically the achilles heel of multi-kinase inhibitors.

The node points to NCT06943755 - zanzalintinib versus everolimus in previously treated unresectable or metastatic neuroendocrine tumors, both pancreatic and extra-pancreatic. Everolimus is the right comparator: it is the established standard second-line agent in NET based on RADIANT-3 (pancreatic NET) and RADIANT-4 (non-functional GI/lung NET). Cabozantinib's recent CABINET trial (NCT03375320) demonstrated PFS benefit over placebo in both NET cohorts but was not a head-to-head against everolimus - so zanzalintinib vs everolimus directly tests displacement of the standard of care, which CABINET did not attempt. Primary endpoint is progression-free survival, the standard regulatory bar in this slow-moving cancer. The bigger picture matters more than any single trial: STELLAR-303 randomized 901 refractory CRC patients 1:1 to zanzalintinib + atezolizumab versus regorafenib and hit on OS (HR 0.80, p=0.0045) [1][5]. STELLAR-304 randomized non-clear cell RCC patients to zanzalintinib + nivolumab versus sunitinib [3]. STELLAR-305 tests zanzalintinib + pembrolizumab versus pembrolizumab alone in PD-L1-positive HNSCC [4]. Combined with the active Exelixis-sponsored solid tumor combination basket (NCT05176483, ~1,300 participants across many cohorts), this is a saturation strategy across virtually every TKI-sensitive solid tumor [10]. The risk in running this many Phase 3 programs simultaneously is that a tolerability signal in one indication can spook investigators and regulators across all of them. The reward is multiple shots on goal.

Our model gives this drug an 18% chance of eventually being approved. That starts from a 48% historical approval rate for Phase 3 drugs in this area, then adjusts based on ten specific facts about the trial and its sponsor. The estimate is pushed up by the trial's open-label design, and pulled down by the sponsor's weak approval track record, limited earlier-phase results, and a randomized trial structure. The remaining factors were close to average and had little effect on the final number.

Efficacy risk: refractory MSS colorectal cancer is the graveyard of oncology, and STELLAR-303 only marginally beat the standard of care in absolute terms. Median OS was 10.9 vs 9.4 months - a 1.5 month absolute delta with HR 0.80 [1]. For context, regorafenib gained ~1.4 months over best supportive care in CORRECT, fruquintinib gained ~2.6 months (7.4 vs 4.8) in FRESCO-2 [11], and TAS-102 gained ~1.8 months in RECOURSE. So STELLAR-303 sits at the lower end of this band against an active comparator (not BSC), which is meaningful - but payers will scrutinize a premium combination price for a 1.5 month gain. Safety risk: multi-kinase inhibitors live and die on tolerability. Cabozantinib's real-world dose reductions are extensive - patients routinely drop from 60mg to 40mg to 20mg. Zanzalintinib's 16-22h half-life [13] is the engineered answer, but the field will need to see whether the Phase 3 discontinuation and dose-reduction rates actually beat cabozantinib in matched indications. Hand-foot syndrome, hypertension, diarrhea, and fatigue are the predictable on-target burdens of VEGFR2 inhibition [10]. Execution risk: running four Phase 3 trials and dozens of combination cohorts at once strains operations. Commercial / patent-cliff risk: the Cabometyx US composition-of-matter patent expires August 14, 2026; later formulation/salt patents and the settled licenses to Teva and Cipla push generic entry to January 1, 2031 [15]. Zanzalintinib has to be on-label and ramping in multiple indications before 2031, AND has to displace cabozantinib in indications where Exelixis already owns the market - every dollar of zanzalintinib revenue in RCC or NET partially cannibalizes cabozantinib revenue. The franchise math only works if zanzalintinib expands into new indications (CRC, HNSCC) faster than it eats existing ones, and if the Merck-partnered LITESPARK-034 program in clear cell RCC delivers a differentiated label [7].

Worth watching closely. STELLAR-303 already de-risked the platform - a Phase 3 OS win in refractory MSS CRC (HR 0.80, median OS 10.9 vs 9.4 months) is a real result and validates the AXL/MER plus VEGFR2 plus checkpoint combination logic, even if the absolute delta is modest [1][2]. The PDUFA date of December 3, 2026 for the mCRC indication is the next hard catalyst [14]. Watch for biomarker subgroup data from STELLAR-303 (PD-L1 expression, AXL/MET expression, MSI status confirmation) - whether the OS benefit is enriched in a subset or applies broadly will determine the commercial story and payer use. The next late-stage signal that matters is STELLAR-304 in non-clear cell RCC, where the bar is meaningful but achievable; positive PFS with a clean HR around 0.6 would establish zanzalintinib as a true cabozantinib successor [3]. The signal that would make this a buy-the-dip moment for Exelixis equity is a STELLAR-305 win in HNSCC, where the I/O + TKI combination story is still wide open and the addressable market is large. Check back at ASCO 2026 and ESMO 2026 for STELLAR-304 and STELLAR-305 maturation, and watch Exelixis 8-Ks for any pre-readout disclosures [12]. Connect this to the Cabometyx revenue trajectory in the Exelixis FY2025 10-K [6] and the 2031 generic-entry date [15] - the company's entire valuation story depends on whether zanzalintinib can extend an oncology franchise that has otherwise been a one-drug operation. The Merck partnership in clear cell RCC is the second-derivative bet: if LITESPARK-034 lands, Exelixis gets a co-developed Merck-distributed product without the commercial overhead [7].